In this project, we explore the use of transformer models to generate particle theory Lagrangians. By treating Lagrangians as complex, rule-based constructs similar to linguistic expressions, we employ transformer architectures —proven in language processing tasks— to model and predict Lagrangians. The ultimate goal of this project is to develop an AI system capable of formulating theoretical explanations to experimental observations, a significant step towards integrating artificial intelligence into the iterative process of theoretical physics.